COMPATIBILITY BETWEEN UMTS 900/1800 AND SYSTEMS OPERATING IN ADJACENT BANDS

Transcription

1 Electronic Communications Committee (ECC) within the European Conference of Postal and Telecommunications Administrations (CEPT) COMPATIBILITY BETWEEN UMTS 900/1800 AND SYSTEMS OPERATING IN ADJACENT BANDS Krakow, March 2007

2 Page 2 1 EXECUTIVE SUMMARY This report deals with the compatibility study between UMTS900/1800 and systems operating in adjacent bands. This report gives the description of the compatibility study methodology, co-existence scenarios, simulation assumptions, and the results for the deployment of UMTS operating in 900 MHz and in 1800 MHz bands taking into account adjacent band systems. Although best effort has been made to provide assumptions and results to encompass the widest range of possible situations, however there might be some country specific cases where different assumptions need to be made. Furthermore it has to be noted that based on the operational experience further analyses may have to be carried out. Based on the interference analysis, the following conclusions can be made: UMTS900 can be deployed in the same geographical area in co-existence with GSM-R as follows: 1) There is a priori no need of an additional guard band between UMTS900 and GSM-R, a carrier separation of 2.8 MHz or more between the UMTS900 carrier and the nearest GSM-R carrier is sufficient without prejudice to provisions in point 2). This conclusion is based on Monte Carlo simulations assumed suitable for typical case. 2) However for some critical cases (e.g. with high located antenna, open and sparsely populated areas served by high power UMTS BS close to the railway tracks, blocking etc, which would lead to assumption of possible direct line of sight coupling) the MCL calculations demonstrate that coordination is needed for a certain range of distances (up to 4 km or more from railway track). 3) It is beneficial to activate GSM-R uplink power control, especially for the train mounted MS, otherwise the impact on UMTS UL capacity could be important when the UMTS network is using the 5 MHz channel adjacent to the GSM-R band. However, it has to be recognized that this is only applicable in low speed areas as elsewhere the use of uplink control in GSM-R will cause significantly increased call drop out rates. 4) In order to protect GSM-R operations, UMTS operators should take care when deploying UMTS in the 900 MHz band, where site engineering measures and/or better* filtering capabilities (providing additional coupling loss in order to match the requirements defined for the critical/specific cases) may be needed in order to install UMTS sites close to the railway track when the UMTS network is using the 5 MHz channel adjacent to the GSM-R band. * Currently, the out-of band interference level is given by 3GPP TS V7.4.0 It has to be noted that this study did not address tunnel coverage. Site sharing, which is expected to improve the coexistence, has not been studied either. When UMTS900 is deployed in the same geographical area in co-existence with PMR/PAMR (CDMA PAMR, TETRA, TAPS) operating at frequencies above 915 MHz, some potential interference from PAMR/PAMR BS to UMTS900 BS could be a problem. In order to protect UMTS900 BS, the utilization of interference mitigation techniques is necessary: i) Reduced PMR/PAMR BS Tx power ii) Spatial separation iii) External filters iv) Guard band The potential interference from UMTS900 to aeronautical DME operating at frequencies above 972 MHz does not represent any difficulty. The frequency range between MHz is not currently used by aeronautical DME but is planned to be used in a near future. Some additional margins may be required for the protection of aeronautical DME operating at frequencies between 960 and 972 MHz, where the required additional margins are dependent on DME carriers and aircraft positions. The studies have shown that the only mitigation techniques, in order to ensure the compatibility between the DME system and UMTS900, that would bring sufficient isolation are: additional filtering and a larger guard band. However these two mitigation techniques are not judged applicable. It has to be noted that the impact of the DME ground station (and FRS if necessary) on the UMTS 900 mobile stations has not been studied in this report and may need additional studies. Therefore, the report suggests that a regulatory solution should be examined. It is necessary that a common approach be used within Europe to ensure the compatibility.

3 Page 3 Further compatibility study will be necessary if this frequency range is to be used by DME systems or future aeronautical systems addressed under WRC Agenda Item 1.6. The compatibility study between UMTS900 and MIDS indicated that an additional margin of 17 db of UMTS900 BS spurious emissions over the frequency range between 1000 MHz and 1206 MHz in reference to 3GPP technical specifications is required for the protection of MIDS terminal receiver. If this additional margin is obtained by the UMTS BS real performance being better than 3GPP technical specifications, no other protection means such as separation distance etc. are required for the protection of MIDS. Potential interference between UMTS1800 and DECT does not appear to be a problem, as the DECT system has a DCA (Dynamic Channel Allocation) mechanism which efficiently avoids an interfered channel except if both systems are deployed indoor. Indeed, although DECT uses DCA, interference analysis shows that in the case of UMTS1800 indoor pico cellular deployment using the frequency channel adjacent to the DECT frequency band, the use of some interference mitigation technique may be necessary to address potential interference to indoor DECT RFP or PP. However, in practice, GSM1800 deployment has demonstrated that no additional interference mitigation techniques are really needed. This statement can be assumed to be extended to the compatibility between UMTS1800 and DECT systems. The analysis indicates that the potential interference between UMTS1800 UE and METSAT Earth Stations should not be a problem. The preliminary interference analysis leads to the conclusion that, with a guard band of 700 khz, the potential interference from Radio microphones to UMTS1800 BS should not be a problem if the radio microphones maximum transmit power is limited to 13 dbm (20 mw) for hand held microphones and 17 dbm (50 mw) for body worn microphones as recommended in ERC Report 63 and ERC/REC 70-03E. It should be noted that the interference analysis between UMTS1800 UE and Fixed Services was not considered in the report. In some European countries, civil/military aeronautical radionavigation system is using the frequency band adjacent to UMTS900, different to the frequency band of civil radionavigation DME, it is also used as safety-of-life application. The frequency plan and the characteristics of the civil/military aeronautical radionavigation system, as well as the interference analysis between UMTS900 and the civil/military aeronautical radionavigation system are not considered in this report.

4 Page 4 Table of contents 1 EXECUTIVE SUMMARY INTRODUCTION COMPATIBILITY STUDY BETWEEN UMTS900 AND SYSTEMS OPERATING IN ADJACENT BANDS SYSTEMS OPERATING IN ADJACENT BANDS COMPATIBILITY STUDY BETWEEN UMTS900 AND GSM-R GSM-R system characteristics Interference analysis based on the comparison of out-of-band emissions between UMTS and GSM Introduction Comparison of UMTS900 and GSM900 out-of-band emissions Analysis summary Interference analysis with MCL approach Introduction Interference analysis results Interference analysis with Monte-Carlo simulations UMTS900 and GSM-R deployment and co-existence scenarios Simulation assumptions Interference analysis method Simulation results Analysis summary Conclusions COMPATIBILITY CONSIDERATION BETWEEN UMTS900 AND PMR/PAMR Characteristics of PMR/PAMR systems CDMA PAMR system characteristics TETRA system characteristics Interference analysis considerations Potential interference between UMTS900 and CDMA PAMR at 915 MHz Potential interference between UMTS900 and TETRA at 915 MHz Conclusions COMPATIBILITY STUDY BETWEEN UMTS900 AND DME DME and UMTS system characteristics Case Study Interference analysis results Analysis of the results Mitigation techniques and mitigation effects Conclusions COMPATIBILITY STUDY BETWEEN UMTS900 AND MIDS System parameters and co-existence scenario Frequency band plan System parameters Propagation model Simulation configuration Interference analysis and simulation results Level of the UMTS900 signal received by the MIDS terminal (out of the MIDS receiving band) Level of the UMTS900 signal received by the MIDS terminal (in the MIDS receiving band) Conclusions CONCLUSIONS COMPATIBILITY STUDY BETWEEN UMTS1800 AND SYSTEMS OPERATING IN ADJACENT BANDS SYSTEMS OPERATING IN ADJACENT BANDS COMPATIBILITY STUDY BETWEEN UMTS1800 AND DECT DECT system characteristics UMTS1800 system characteristics Interference analysis between UMTS1800 and DECT Interference analysis and simulation results Conclusions COMPATIBILITY CONSIDERATION BETWEEN UMTS1800 AND METSAT Main characteristics of METSAT Interference analysis considerations...59

5 Page Conclusions COMPATIBILITY CONSIDERATION BETWEEN UMTS1800 AND RADIO MICROPHONES Main characteristics of Radio Microphones Interference analysis Conclusions COMPATIBILITY STUDY BETWEEN UMTS1800 AND FIXED SERVICES CONCLUSIONS REFERENCES...63 ANNEX 1 - GSM900 AND UMTS900 ACLR PROFILES...64 ANNEX 2 - INTERFERENCE ANALYSIS CALCULATION WITH MCL APPROACH FOR THE CO- EXISTENCE BETWEEN UMTS900 AND GSM-R...66 ANNEX 3 - ABBREVIATIONS...73

6 Page 6 2 INTRODUCTION UMTS networks have been widely deployed in the frequency band MHz/ MHz), however, there are still sparsely populated and remote areas where there are difficulties to provide IMT-2000/UMTS services in a costefficient way. UMTS deployment in the 900 MHz band can facilitate the provision of the expected IMT-2000/UMTS services to users in those areas. The main interest for European mobile operators to deploy UMTS in the 900 MHz band is the larger coverage compared to UMTS in the 2000 MHz band. UMTS900 offers a considerably more cost effective solution for providing UMTS services in rural area with low population density. The total bandwidth of the 1800 MHz frequency band is 2x 75 MHz. In some countries, the 1800 MHz band is not totally used by GSM systems, especially in low population density rural areas. Part of the 1800 MHz band may become a complementary band for deploying UMTS, where the interest for mobile operators to deploy UMTS comes also from the fact that it is easy to share the same GSM1800 radio sites by UMTS systems operating in 1800 MHz band. The 900 MHz band and 1800 MHz band have been allocated to GSM systems in Europe and they are widely used. As deployment of UMTS (UTRA-FDD) systems in the 900 MHz band and 1800 MHz band does not mean the replacement of GSM systems by UMTS, good compatibility between UMTS and GSM in the 900 MHz and 1800 MHz bands is important and necessary. ECC Report 82 deals with the compatibility study for UMTS deployed in the GSM900 and GSM1800 frequency bands. The deployment scenarios of UMTS900/1800 and potential interference between UMTS and GSM operating in adjacent channels have been described in ECC Report 82. European frequency allocation tables indicate that several systems are using frequency bands adjacent to UMTS900/1800 (GSM900/1800) systems, and several ERC and ECC Reports have been developed on the compatibility between GSM900/1800 and systems operating in adjacent bands. The intention of this report is to deal with the compatibility study between UMTS900/1800 and systems operating in adjacent bands. This report gives the relevant parameters of systems operating in adjacent bands of UMTS900/1800, which are needed in interference studies. The interference problems are investigated by both deterministic and statistical approaches. Some scenarios are studied with detailed simulations and analysis, for example the interference scenarios between UMTS900 and GSM-R and the interference scenarios between UMTS1800 and DECT, whereas the potential interference analysis for several other cases are considered and derived from the existing ERC and ECC Reports for GSM900/1800. In this report, chapter 3 is dedicated to the compatibility study between UMTS900 and systems operating in its adjacent bands. The compatibility study between UMTS1800 and the adjacent band systems is described in chapter 4. 3 COMPATIBILITY STUDY BETWEEN UMTS900 AND SYSTEMS OPERATING IN ADJACENT BANDS 3.1 Systems operating in adjacent bands All systems operating in bands adjacent to UMTS900 and addressed in this report are summarized in table 3-1 below. Frequency (MHz) System Note GSM-R (UL) GSM900 (UL) Including E-GSM and P-GSM UMTS900 (UL) PMR/PAMR (DL) GSM-R (DL) GSM900 (DL) UMTS900 (DL) Including E-GSM and P-GSM Aeronautical DME/TACAN Radionavigation MIDS (Military / NATO) Communication systems Table 3-1: Systems operating in adjacent bands of UMTS900

7 Page 7 The sharing studies between UMTS900 and the following systems operating in adjacent bands were considered: 1) GSM-R 2) PMR/PAMR (e.g. TETRA, TAPS, CDMA) 3) DME 4) MIDS The interference analysis between UMTS900 and GSM-R is described in section 3.2. Section 3.3 gives a brief description of the interference analysis between UMTS900 and PMR/PAMR. The co-existence scenario and the interference analysis between UMTS900 and aeronautical system DME (Distance Measuring Equipment) is described in section 3.4. The interference analysis between UMTS900 and MIDS is described in section 3.5. The conclusion is given in section 3.6. At the same time aeronautical radionavigation systems are operating in the frequency band MHz in some countries (see Radio Regulations). Compatibility studies with these systems were not considered in this Report. 3.2 Compatibility study between UMTS900 and GSM-R The UMTS900 frequency band is arranged as: Uplink (UE transmit, BS receive): MHz Downlink (BS transmit, UE receive): MHz Carrier separation: 5 MHz The GSM-R frequency band is arranged as: Uplink (MS transmit, BS receive): MHz Downlink (BS transmit, MS receive): MHz Carrier separation: 200 khz The frequency band plans for GSM-R and UMTS are shown in figure MHz 880 MHz 915 MHz RGSM-UL UL UMTS-UL RGSM-DL UMTS-DL 921 MHz 925 MHz 960 MHz Figure 3-1: Frequency band plan for GSM-R and UMTS in 900 MHz band GSM-R system characteristics Details of the GSM-R RF performance and system parameters can be found in 3GPP technical specification TS [6]. See also [22]. The main GSM-R system characteristics are summarized in tables 3-2, 3-3, 3-4, and 3-5.

8 Page 8 GSM-R Frequency band (UL) (MHz) Frequency band (DL) (MHz) Carrier separation (khz) 200 Modulation GMSK BS-MS MCL (db) 60 (urban area) 70 (rural area) Typical cell range (km) 8 BS Hand portable MS Train Mounted MS Maximum Tx power (W) Thermal noise (dbm) Noise figure (db) Noise floor (dbm) Receiver sensitivity (dbm) Receiver protection ratio (db) Antenna height (m) 20 (Urban) 45 (Rural) Antenna gain (dbi) Feeder loss (db) Spectrum mask and spurious emissions 3GPP TS GPP TS Table 3-2: Main GSM-R system parameters BS Tx power (khz) (khz) (khz) (khz) (dbm) < (khz) < (khz) < (khz) (khz) 43 +0, * , * , * , * , * , * NOTE: * For equipment supporting 8-PSK, the requirement for 8-PSK modulation is -56 db. Table 3-3: Spectrum mask of GSM-R BTS* *Note: The values given in this table are the maximum allowed level (db) relative to a measurement in 30 khz on the carrier as defined in 3GPP TS [6].

9 Page 9 BS MS General requirement -36 dbm* -36 dbm* Co-siting with GSM dbm/100 khz Table 3-4: Spurious emission of GSM-R MS * measurement band depends on the carrier separation, which is defined in TS [6]. Frequency GSM-R band other MS small MS BTS dbµv dbm dbµv dbm dbµv dbm (emf) (emf) (emf) In-band 600 khz f-f o < 800 khz khz f-f o < 1,6 MHz ,6 MHz f-f o < 3 MHz MHz f-f o out-of-band (a) (b) (c) (d) Table 3-5: Blocking characteristics of GSM-R The cases (a), (b), (c), (d) are defined in 3GPP TS [6] Interference analysis based on the comparison of out-of-band emissions between UMTS and GSM Introduction GSM has been deployed over many years and GSM-R has been deployed in some European countries, and no problem of interference from GSM emissions into GSM-R has been raised so far. This section deals with the comparison of out-ofband emissions between GSM and UMTS. The comparison of emission masks is very helpful in evaluating the potential interference from UMTS900 to GSM-R Comparison of UMTS900 and GSM900 out-of-band emissions Definition of out-of-band emissions Out-of-band emissions are defined in the GSM900 and UMTS900 technical specifications. The ACLR (Adjacent Channel power Leakage Ratio) can be obtained by the integration of the spectrum mask over 200 khz, the ACLR profiles of GSM900 and UMTS900 are given in Annex 1. Assumptions In practical GSM deployment, a sector of a GSM site has several emitters (TRX) and thus, is using several 200 khz GSM carriers over a band of 5 MHz, in order to meet the capacity requirement. It is therefore intended to derive the GSM900 out-of-band emissions with several GSM channels being aggregated. Current GSM networks use a 1x3 re-use scheme for TCH channels; in other words, each TRX is using one carrier randomly chosen among a list of three carriers. For instance, a tri-sector GSM900 base-station using 3 TRX is using 3 GSM frequency carriers. As a consequence, it can be assumed that a GSM sector is using three carriers over 5 MHz. The case of three GSM carriers will be taken into account in the comparison of out-of-band emissions between GSM and UMTS.

10 Page 10 Figure 3-2 below shows the spectral occupancy for the case where the GSM deployment is using a 1x3 frequency re-use scheme and three TRX are implemented in a sector with three 200 khz carriers. It is assumed that the GSM carriers are equally distributed. A worst case scenario would be where all the GSM carriers are located close to the GSM-R allocation, whereas the best scenario in terms of interference would be to have the three carriers as far as possible from the GSM-R allocation. 200 khz guard band 5 MHz GSM-R carrier GSM900 carrier GSM900 carrier GSM900 carrier GSM-R Allocation GSM Allocation Figure 3-2: Spectral occupancy of an E-GSM sector over 5 MHz Comparison of BS out-of-band emissions Out-of-band Emissions - Base Stations dbm (in 200 khz) UMTS900 (50 %) 40 GSM UMTS900 (Pmin) 30 UMTS900 (Pmax) GSM-R/UMTS carrier separation 5 2,6 2,8 3,0 3,2 3,4 3,6 3,8 4,0 4,2 4,4 4,6 4, ,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 2,2 2,4 GSM-R/GSM carrier separation Carrier Separation (MHz) Figure 3-3: Comparison of the BS out-of-band emissions Figure 3-3 gives the out-of-band emissions when three GSM channels are deployed as shown in figure 3-2. The GSM outof-band profile is compared with the UMTS spectrum mask. The transmitting power of both GSM and UMTS BSs are fixed as 43 dbm.

11 Page 11 Figure 3-3 shows the out-of-band emissions when the UMTS sector is transmitting at Pmin, Pmax and also when the cell load is at 50 %, where Pmin is the transmitting power of common channels, and the transmitter power at cell load of 50% is calculated by the addition of common channel powers and the transmitting power for traffic channels at 50% cell load. It should be noted that in the GSM mask, after 2 MHz carrier separation, it enters the spurious emission domain. GSM BS BCCH channel s maximum transmitting power does not depend on the traffic load in the cell and is fixed at its maximum power. On other traffic channels there may be a reduction in mean transmitted power when power control is used, which can be further reduced by the use of DTX. UMTS BS transmitting power is dependent on the traffic load, where usually 10% of the BS power is allocated to the common channel (Pilot, Synch, etc) and the rest of the BS Tx power is allocated to the traffic, depending on the cell load. When the traffic load is zero, then the Pmin = 10% of the BS transmitting power, when traffic load is 50%, the BS Tx Power = (10% +50%x90%) x Maximum TX BS Power; thus TxPower = 55% x Maximum TX BS Power or TxPower = 43 dbm 2.6 db = 40.4 dbm. It should be noted that 50% of cell load is the reference cell load in UMTS network design, whereas in rural areas the reference cell load could be lower than 30% in a coverage driven design. When the cell load is 100%, then the UMTS BS will transmit at its maximum power Pmax. It should be noted that 3GPP technical specification TS and ETSI specification TS defined only the UMTS BS spectrum mask and out-of-band emission limits at the maximum transmitting power. The out-of-band emissions at reduced transmitting power Pmin and P(50%) are calculated under the assumption that the UMTS BS spectrum mask (ACLR) is the same as that at Pmax as defined in ERC Report 68. It can be seen from the curves in figure 3-3 that at 2.8 MHz carrier separation the out-of-band emission of UMTS is lower than that of GSM900 at 400 khz carrier separation, and at 3 MHz carrier separation it is above the GSM900 BS out-ofband emission at 600 khz. Comparison of Terminal out-of-band emissions Figure 3-4 below gives the comparison of out-of-band emissions between UMTS UE and GSM MS, where the cumulative effect of three GSM channels as shown in Figure 3-3 was taken into account in the calculation of GSM MS out-of-band emissions. It should be noted that the variation in locations of the GSM MS relative to the UMTS UE are ignored as this will average out. The GSM900 terminal power is fixed at 33 dbm and the UMTS900 terminal power at 21 dbm. Out-of-band Emissions - Terminals GSM UMTS dbm (in 200 khz) 10 0 GSM-R/UMTS carrier separation 2,6 2,8 3,0 3,2 3,4 3,6 3,8 4,0 4,2 4,4 4,6 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 2,2 GSM-R/GSM carrier separation Carrier Separation (MHz) Figure 3-4: Comparison of the Terminal out-of-band emissions (Pmax) Without taking into account the power control effect, figure 3-4 shows that at 2.8 MHz carrier separation UMTS UE outof-band emission is at the same level as GSM900 MS, at 3.0 MHz carrier separation it is higher, but still below the out-ofband emission of GSM900 MS at 400 khz carrier separation.

12 Page 12 Implementing power control in GSM900 and UMTS900 terminals helps to reduce emission levels drastically. ECC Report 82 (Compatibility Study For UMTS Operating within the GSM 900 and GSM 1800 Frequency Bands) provides the CDF (Cumulative Distribution Function) of UMTS900 UE transmit power. For outdoor UE, it should be noted that 90% of terminals transmit at a power level lower than -23 dbm and 50 % at a power level lower than -32 dbm. It should also be noted that GSM MS power control is much less fast and less efficient compared to UMTS UE power control Analysis summary The comparison of out-of-band emissions between GSM900 and UMTS900 shows that the UMTS900 and GSM900 out-ofband emissions do not present significant difference, which means that UMTS900 should not a priori cause more interference than GSM900. GSM-R has been deployed in many European countries, although experience with uncoordinated use adjacent to the lowest E-GSM frequencies is limited. Based on the comparison of out-of-band emissions between GSM900 and UMTS900, a 2.8 MHz carrier separation between UMTS carrier and the nearest GSM-R carrier is a priori sufficient to ensure the protection of GSM-R based on the above approach Interference analysis with MCL approach Introduction This section deals with the interference analysis between UMTS900 and GSM-R using an MCL (Minimum Coupling Loss) approach. The interference analysis described in this section covers both in-band blocking and out-of-band blocking Interference analysis results Out-of-band emissions Using the out-of-band emission figures (UMTS BS Pmax out-of-band emission curve) described in Section the exclusion distances have been calculated with an MCL approach for the protection of GSM-R Interference analysis results UMTS BS to GSM-R MS The calculations are provided in Annex 2, Part A. Case 1 (based on the Hata-Okumura model) shows that the interference distances for a GSM-R MS operating at minimum GSM-R network design signal level are 3.6 km for speech and 4.4 km for data in the highest GSM-R channel; and 1.8 km for speech and 2.2 km for data in the fourth channel. Even moving to beyond the fourth channel will give interference distances of 1.5 km and 1.9 km for speech and data respectively. This problem can be reduced by the addition of filters in the UMTS BS, however this is unlikely to solve the problem for the highest GSM-R channel unless the filter response is very sharp. It should be noted that no account has been taken of the effect of multiple UMTS transmitters in these calculations. Two alternative calculations are given as additional examples in Annex 2. On Case 1bis the assumptions for the calculation are considered as more conservative in order to address critical situations (including contribution of direct line of sight). One other example of calculations is provided Part B of Annex 2 for the cases where the GSM-R network is noise-limited and interference-limited. The blocking performance of GSM-R mobiles is defined in EN For the GSM-R MS it is defined as -38 dbm for khz carrier separation and this figure has been used below. However, it should be noted that when the difference between the centre of the GSM-R and UMTS channel is set at 2.8 MHz (band edge separation of 200 khz) then the interference will be in-band and a worse blocking performance will be experienced. As shown in Annex 2 - Part A, Case 2, this equates to a distance of 420 m for the high power UMTS BS. If the fading margin is ignored, the blocking distance rises to 830 m.

13 Page 13 UMTS BS to GSM-R BTS Annex 2 - Part A, Case 3 covers the blocking of the GSM-R BTS by the high power UMTS900 BTS. This demonstrates that blocking will occur at a distance of 664 m, increasing to 1.3 km if the fading margin is ignored. Annex 2 - Part A, Case 4 shows that even a low power medium range UMTS BS in a micro-cellular deployment will cause blocking at distances of 175 m, increasing to 320m if the fading margin is ignored. Blocking of the GSM-R BTS could be reduced by applying filters at the GSM-R BTS. The definition of receiver blocking is the effect of a strong out-of-band signal, present at the input of the receiver, on the receiver s ability to detect an in-band wanted signal. Thus, the blocking signal reduces the specified receiver sensitivity by a certain number of dbs. In the case of GSM-R BS receive/umts-bs transmitting, the blocking effects from a UMTS BS have to be compared with what would occur from a GSM BTS. Noting that the height of the antennas, tilt, gain, and sector aperture will be the same for GSM and UMTS BS, two elements need to be considered: - The max EIRP from the interferer and the resultant interfering level at the victim BS, including selectivity properties of the receiver; - The occurrence probability of blocking issues with regard to 2G/3G air interface refarming. As gains are the same for both UMTS and GSM BS, we will just compare transmitting power of each technology. When UMTS maximum transmit power is 43 dbm/3.84 MHz, GSM BS transmit power is 43 dbm/200 khz with a number of simultaneous channels transmitting, dependent on the size of band allocated to one operator and the frequency reuse factor. It has to be noted that UMTS downlink is power controlled in order to reduce the transmitting power to between 33 and 43 dbm/3.84 MHz. GSM BS transmits at radio frequency channels without power control such as BCCH (Broadcast Common Channel) channels using the full power of 43 dbm/200 khz. Thus from a transmitting power point of view, GSM BS could cause more severe blocking to GSM-R than UMTS BS. Concerning occurrence probability, using the same cell sites, a GSM network with 4x12 frequency reuse factor for radio frequency channels without power control and frequency hopping, 4x12 or 1x3 for TCH with or without frequency hopping and power control, the probability to have a GSM BTS transmitting in close geographical vicinity of a GSM-R base receiver will be the same as for a UMTS network with frequency re-use 1 scheme Analysis summary Generally it is considered that the MCL method for interference analysis is the worst case where no system outage is accepted, and consequently the results are usually pessimistic. In the interference analysis with MCL approach presented above, the minimum allowed signal level used was the network design objective level of -98 dbm, and not the GSM-R MS receiver sensitivity level, which allows for a small probability of outage at the limits of coverage. From the interference analysis results shown above with MCL approach, it is apparent that considerable interference at distances of greater than 2 km will be caused to GSM-R systems if the lower UMTS channels are used. This result will also be applicable to lower power UMTS BS, although the interference distance is reduced to 170 m. A reduction in the effects of out-of-band emissions can be achieved by applying filtering to the UMTS BS, but it is considered that this will still require a suitable guard band. The effect of blocking is more significant and requires that no high power UMTS900 BS is placed closer than 660 m to the railway without coordination. Even low power micro-cells will need to be placed at a distance of at least 170 m, which will rule out their deployment inside railway station areas.

14 Page Interference analysis with Monte-Carlo simulations UMTS900 and GSM-R deployment and co-existence scenarios GSM-R deployment scenario GSM-R networks offer a linear coverage of railway lines with bi-sector radio sites installed along the railway, as shown in figure 3-5. The main system characteristics and network parameters are summarized in table 3-2. Figure 3-5: GSM-R deployment scenario Two major characteristics of GSM-R coverage are: 1) Linear coverage; 2) High quality coverage (95% space and time availability). In Europe, most GSM-R networks are designed with a BS antenna height of about 30 m, and cell range is around 5-6 km. The assumption of BS antenna height at 45 m and cell range at 8 km represents the worst case scenario for the sharing study. There are two types of GSM-R MS as described in table 3-2: 2W handset MS and 8W train mounted MS. As shown in figure 3-6 below, the GSM-R 8W train mounted MS is the MS that is located inside the train, connected to the external MS antenna mounted on the roof top of the train. BTS Figure 3-6: Connection between train mounted antenna and MS situated inside of the train

15 Page UMTS900 deployment scenario The main objective of UMTS deployment in the 900 MHz band is for coverage extension, but in urban areas the deployment of UMTS in the 900 MHz band can also improve tremendously the indoor coverage quality. In rural areas the deployment of UMTS in the 900 MHz band allows mobile network operators to offer 3G services at lower cost. UMTS Inter-site distance 3*R Cell radius R Cellrange 2*R Figure 3-7: UMTS network layout The typical UMTS900 deployment scenario considered in the sharing study with GSM-R is the rural area deployment with cell range 2*R=5000 m, where the network layout is shown in figure Co-existence between UMTS900 and GSM-R Based on the GSM-R and UMTS deployment scenarios described above, simulations were performed based on the coexistence scenario shown in figure 3-8. As shown in figure 3-5, GSM-R BS sites are placed along the railway, where the average distance between GSM-R BS radio site and the railway is 20 m. The separation distance between the railway line and UMTS sites is represented by d0. Table 3-6 below gives three typical distance shift r=d0/4330 m and the separation distances between railway line and the nearest UMTS sites, where the distance 4330 m is obtained from 2*R*cos(30 )=5000*cos(30 )=4330 m, as shown in figure 3-8. Distance shift r Separation distance (r=d0/4330) d0 (m) Table 3-6: Distance between railway line and UMTS sites The simulation was done with a quasi-static Monte-Carlo simulator. UMTS UEs are randomly distributed within the UMTS900 coverage area, the reference UMTS network uplink and downlink capacities are simulated without the presence of GSM-R network. The UMTS uplink capacity is obtained with the threshold of 6 db noise rise, corresponding to 75% cell load. Downlink capacity is simulated with the threshold of 5% outage.

16 Page 16 2 x 104 d Figure 3-8. UMTS900 and GSM-R Co-existence scenario x 10 4 GSM-R MS are uniformly distributed on the railway line, GSM-R downlink channel is considered as radio frequency channels without power control, but power control on uplink is activated in the simulations. The reference GSM-R performance is the uplink and downlink outages without interference from UMTS. It should be noted that this co-existence scenario is valid for rural environment. It is important to note that these simulations didn t consider dynamic behavior of GSM-R (e.g. for the case of deployment of high speed trains) and UMTS900 systems. Additional studies for those cases may be needed on a national basis, based on practical experience Simulation assumptions Simulation assumptions are summarized in table 3-7. These assumptions are similar to those used in the sharing study between UMTS900 and GSM900 described in ECC Report 82[1]. Scenario Simulation cases UMTS(macro) - GSM-R(macro) in rural area in uncoordinated operation Two simulation cases : UMTS uplink as victim and GSM-R DL as victim 1) GSM-R Downlink -GSM-R (radio frequency carrier without power control) as victim for train mounted GSM-R MS 2) UMTS Uplink - WCDMA UL as victim (Simulate GSM system, then add UMTS users until the total noise rise hits 6 db) - GSM-R uplink power control is activated -No frequency hopping for GSM-R -Run simulations with various ACIRs between UMTS carrier and the nearest GSM-R carrier for various space separations between UMTS radio site and railway (d0).

17 Page 17 Network layout As shown in figure 3-8 above - Rural environment - 3-sector configuration for UMTS network - bisector configuration for GSM-R network - GSM-R frequency reuse 6, as shown in figure sites (i.e., 57 cells (sectors)) with wrap-around for UMTS - UMTS Cell radius R=2500m, cell range 2R=5000m, inter-site distance 3R= 7500 m (as shown in figure 3-7) -GSM-R cell range : 8 km -Distance between GSM-R radio site and railway: 20 m System parameters Services Propagation model WCDMA BS antenna gain with cable loss included = 15dBi BS antenna height H bs =45 m; UE antenna height H ms =1.5 m BS-UE MCL=80 db BS antenna(65 horizontal opening) radiation pattern is referred to 3GPP TR V6.0.0 ( ), Section A.3 (Annex) UE antenna gain 0 dbi (omni-directional pattern) GSM-R BS antenna gain with cable loss included = 15dBi BS antenna height H bs =45 m; MS antenna height H ms =4.5 m (train mounted MS) BS-MS MCL=70.5 db BS antenna(65 horizontal opening) radiation pattern is referred to 3GPP TR V6.0.0 ( ), Section A.3 (Annex) Train mounted MS antenna gain 2 dbi (omni-directional pattern) WCDMA 8 kbps Speech (chip rate: 3.84 Mcps) o Eb/Nt target (downlink): 7.9 db o Eb/Nt target (uplink): 6.1 db GSM-R Speech SINR target (downlink): 9 db for speech and 12 db for data SINR target (uplink): 6 db WCDMA and GSM-R Log_normal_Fading = 10 db Rural area propagation model (Hata model) L (R)= log f 13.82log(H b )+[ log(H b )]logr 4.78(Log f) log f a(hm) Hb is BS antenna height above ground in m, f is frequency in MHz, R is distance in km, Hm is the MS antenna height in m. a (Hm) = [1.1*log(f) - 0.7]*Hm - [1.56*log(f) - 0.8] With Hb=45m, Hm=1.5m, f=920 MHz, the propagation model for UMTS UE is simplified as L1( R) =34.1*log(R) With Hb=45m, Hm=4.5m, f=920 MHz, the propagation model for GSM-R MS is simplified as L2( R) =34.1*log(R) The path loss from a transmitter antenna connector to a receiver antenna connector (including both antenna gains and cable losses) will be determined by: Path_Loss = max (L(R) + Log_normal_Fading - G_Tx G_Rx, Free_Space_Loss + Log_normal_Fading - G_Tx G_Rx, MCL) where G_Tx is the transmitter antenna gain in the direction toward the receiver antenna, which takes into account the transmitter antenna pattern and cable loss,

18 Page 18 Cell selection SIR calculation Power Control assumption Capacity ACIR G_Rx is the receiver antenna gain in the direction toward the transmitter antenna, which takes into account the receiver antenna pattern and cable loss, Log_normal_Fading is the shadowing fade following the log-normal distribution. WCDMA As per 3GPP TR GSM-R As for WCDMA in 3GPP TR , but with only one link selected at random within a 3 db handover margin WCDMA As per TR , except for the following changes: Interference contributions from GSM TRXs or MSs are added to the total noise-plus-interference. Processing gain is changed to 26.8 db for 8 kbps Thermal noise level is -103 dbm for uplink Thermal noise level is raised to -96 dbm for downlink GSM-R Total noise-plus-interference is sum of thermal noise, GSM-R co-channel, and WCDMA interference. Cells are synchronised on a time slot basis. Adjacent channel GSM interference is neglected. Noise floor (downlink): -111 dbm Noise floor (uplink): -113 dbm WCDMA As per 3GPP TR dbm terminals Maximum BS power: 43 dbm Maximum power per DL traffic channel: 30 dbm Minimum BS power per user: 15 dbm. Minimum UE power: 50 dbm. Total CCH power: 33 dbm GSM-R Stabilization algorithm same as for WCDMA (C/I based) with a margin of 5 db added to the SIR target. Maximum power (TRx): 43 dbm Minimum power (TRx): 10 dbm (radio frequency carrier with power control) Maximum power (MS): 39 dbm Minimum power (MS): 5 dbm WCDMA Capacity loss versus ACIR as per 3GPP TR GSM-R WCDMA to GSM-R GSM-R Load to maximum number of users and observe change in outage (i.e., 0.5 db less than SINR target) As per spectrum masks defined in TS , TS (applying the appropriate measurement BW correction), unless capacity loss is found to be significant. ACIR( f ) = C( f 0 ) + m ( f f 0 ) (db) GSM-R BTS to WCDMA UE: Consider 3GPP TS45005 GSM BTS transmitter emission mask for 900 band and WCDMA UE receiver selectivity slope, m = 0.8 db / 200 khz GSM-R MS to WCDMA BS: Consider 3GPP TS45005 GSM-R MS transmitter emission mask for 900 band and WCDMA BS receiver characteristics, m = 0.5 db / 200 khz Table 3-7: Summary of simulation parameters for the co-existence between UMTS900 and GSM-R

19 Page MHz f6 f5 f4 f3 f2 f1 f0 f6 f4 f2 f5 f1 f3 Figure 3-9: GSM-R frequency reuse GSM-R carriers arrangement relative to UMTS carrier and the GSM-R frequency re-use plan are given in figure Interference analysis method Interference between UMTS operating in the 900 MHz band and GSM-R was analyzed with the method of Monte-Carlo simulations. The same simulation tools used for the sharing study between UMTS900 and GSM900 as described in ECC Report 82[2] was used for performing simulations for the co-existence between UMTS900 and GSM-R based on the coexistence scenario described above. The objective of Monte-Carlo simulations is to determine the interference between UMTS900 and GSM-R at different carrier separations and at different space separations between the railway line and UMTS sites. ACIR values for UMTS DL/UL as victims and for the GSM system (used in this study as GSM-R) DL/UL as victims at carrier separations of 2.8 MHz and 4.8 MHz are calculated and described in the ECC Report 82 [1]. They are summarized in tables 3-8 and 3-9. Carrier separation 2.8 MHz 4.8 MHz UMTS UL as victim UMTS DL as victim UMTS UL as victim UMTS DL as victim ACIR (db) > 47.4 > 30.5 Table 3-8: ACIR for UMTS UL/DL as victim when being interfered by GSM-R UL/DL Carrier separation 2.8 MHz 4.8 MHz GSM-R UL as victim GSM-R DL as victim GSM-R UL as victim GSM-R DL as victim ACIR (db) Table 3-9: ACIR for GSM-R UL/DL as victim when being interfered by UMTS UL/DL Two simulation cases were studied: 1) GSM-R DL outage degradation based on C/I threshold due to interference from UMTS BS. 2) UMTS uplink capacity loss due to interference from GSM-R 8 W train mounted MS.

20 Page Simulation results Probability of GSM-R DL outage (%) The simulated GSM-R DL outage with speech service C/I=9 db without interference from UMTS based on the frequency reuse plan given in figure 3-9 is nearly zero. The probability of GSM-R DL outage (C/I=9 db) as a function of ACIR between UMTS carrier and the nearest GSM-R carrier for different space separation distances between UMTS BS site and railway line (d0 as indicated on figure 3-8 and distance shift r in table 3-6) was simulated. The simulation curves for different distance offsets are plotted in figure R-GSM DL outage probability due to interference from UMTS DL R-GSM DL outage probability (%) ACIR (db) Qualcomm (r=0) Nortel (r=0) Qualcomm (r=0.5) Nortel (r=0.5) Qualcomm (r=1) Nortel (r=1) Figure Probability of GSM-R DL Outage (%) (C/I=9 db) As shown in figure 3-10, at the operating point ACIR=50 db which corresponds to a carrier separation of 2.8 MHz between the UMTS carrier and the nearest GSM-R carrier, the GSM-R DL outage probability is smaller than 0.15% for the worst case when UMTS sites are co-aligned with GSM-R railway sites. When UMTS sites are not on the railway track, the interference is even smaller. When the railway and UMTS sites are separated by 2165 m, GSM-R DL outage probability is smaller than 0.045%, and when the separation distance is 4330 m, the GSM-R DL outage probability is below 0.015%. From the simulation results, it can be considered that the interference from UMTS DL to GSM-R DL train-mounted MS is under acceptable level, and that no additional guard band is required for the protection of GSM-R DL. Thus UMTS can be deployed in the same geographical area with a carrier separation of 2.8 MHz between the UMTS carrier and the nearest GSM-R carrier. For the co-existence scenario between UMTS900 and GSM-R described in section and simulation assumptions described in section , simulations on the interference from UMTS900 DL to GSM-R DL reception of train mounted MS have been performed with CEPT simulation tool SEAMCAT 3 (Version ) for the thresholds of C/I=9 db and C/I=12 db, the simulation results are presented in figure 3-10a. The considered carrier separation between UMTS carrier and the nearest GSM-R carrier is 2.8 MHz. The SEAMCAT scenario file for this study is attached to this report (can be found at the website next to the downloadable file of this report).

21 Page Probability of simulated C/I on GSM-R DL P(C/I<=9 db) (%) P(C/I<=12dB) (%) Probability (%) Separation distance between Railway line and UMTS sites Figure 3-10a: Probability of simulated C/I on GSM-R DL due to interference from UMTS900 BS The simulation results given in the figure 3-10a simulated with Seamcat show that for the probability of C/I 9 db is smaller than 0.02%, and that of C/I 12 db is smaller than 0.1%, which is below the required 0.5%. Additional studies have been carried out in order to assess the worst-case situations corresponding to the GSM-R devices at the cell-edge. The considered carrier separation between UMTS and nearest GSM is 2.8 MHz. Two distinct scenarios have been investigated. The first one is based on the scenario described in section and , in which the BTS antenna height is fixed at 45 m and the GSM-R cell range is 8 km. The distance between the GSM-R BTS and the GSM-R train mounted MS (Hms=4.5m) is randomly drawn between 7 and 8 km for the different snapshots, in order to simulate the situation in which the GSM-R train mounted MS is far from the serving base station and is at the cell edge. The figure below shows this worst case GSM-R configuration simulated in Seamcat. GSM-R devices randomly drawn 0 km 7 km 8 km Figure 3-10b: Position of GSM-R train mounted MS relative to the serving BS

22 Page 22 An additional scenario was also considered; the antenna height is reduced to 25 m, as well as the cell range to 5 km. For this case, the GSM-R train mounted MSs are randomly distributed between 4 and 5 km in order to simulate the situation in which the GSM-R train mounted MS is far from the serving base station and is close to the cell edge. For both scenarios, the distance between UMTS sites and GSM-R railway is 0, i.e. the UMTS BS sites are placed along the railway. (see figure 3-8 in section ). The following table gives the simulated probability of interference from UMTS900 BS to the GSM-R train mounted MS when considering a C/I ratio of respectively 9 db and 12 db. Antenna height Position of GSM-R Users 45 m 7-8 km 0 25 m 4-5 km 0 Separation distance between UMTS BS and Railway C/I (GSM-R DL) Probability of interference 12 db 0.25 % 9 db 0.06 % 12 db 0.14 % 9 db 0.04 % Table 3-9a: Probability of interference from UMTS900 BS to GSM-R train mounted MS It can be seen that even for the C/I of 12 db, the probability of interference is smaller than 0.25 %, which is below the required 0.5% outage level UMTS UL Capacity Loss (%) Simulation results of UMTS uplink capacity loss (%) as a function of ACIR between the UMTS carrier and the nearest GSM-R carrier for different distance offsets are plotted in figure As shown in figure 3-11, at the operating point ACIR=43.1 db (which corresponds to a carrier separation of 2.8 MHz between the UMTS carrier and the nearest GSM-R carrier), the UMTS UL capacity loss due to interference from GSM-R train mounted MS is smaller than 1.5%, for the worst case when UMTS sites are co-aligned with GSM-R railway (i.e. with distance offset r=0). When UMTS sites are not co-aligned with the railway track, the interference is even smaller. When the railway and UMTS sites are separated by 2165 m or 4330 m, UMTS UL capacity loss is smaller than 0.3%. It should be noted that the UMTS uplink capacity loss is simulated for the whole UMTS network, some of the UMTS cells are more impacted by the interference from GSM-R UL than other cells. Cell no. 31 (one of the nearest cell to railway track) in the network layout shown in figure 3-8 was found to be the worst cell. The UMTS uplink cell capacity loss for a single cell can not be easily simulated, but it can be calculated based on the received noise rise recorded for a specific cell; using the N-pole capacity formula with 75% reference cell load, the uplink cell capacity loss can be estimated. The obtained UL cell capacity loss for cell no.31 is 1.95%, for the case when UMTS sites are co-aligned with the railway track (r=0). UMTS UL Capacity Loss (%) due to interference from R-GSM UL Capacity Loss (%) ACIR (db) Nortel (r=0) Qualcomm (r=0) Qualcomm (r=0.5) Qualcomm (r=1) Figure 3-11: UMTS UL capacity loss (%) due to interference from GSM-R UL

23 Page 23 As described in the above section, GSM-R UL power control is activated in the simulations. The simulated GSM-R train mounted MS Tx power distribution is plotted in figure It can be seen that only 0.5% of MS transmit at maximum power of 39 dbm. R-GSM MS Tx Power Distribution (TxP<=P0) Probability (TxP<=P0) (%) Tx Power P0 (dbm) Figure 3-12: Simulated GSM-R train mounted MS Tx power distribution If GSM-R UL power control is not activated, all GSM-R train mounted MS will transmit at maximum power of 39 dbm. For the case of UMTS sites being placed aligned with the railway track or close to the railway track, the impact on UMTS UL capacity loss due to interference from GSM-R train mounted MS could become much more important. In that case, UMTS operators may need to take care of that problem by using site engineering solutions to reduce the potential interference from GSM-R UL when the UMTS network is using the 5 MHz channel adjacent to GSM-R band Analysis summary Under the assumptions described above, the Monte-Carlo simulation results show that the impact on GSM-R DL by the potential interference from UMTS DL is very low, and a carrier separation of 2.8 MHz between the UMTS900 carrier and the nearest GSM-R carrier should be enough. When GSM-R UL power is used, the simulation results indicate the UMTS900 network capacity loss is below 5% even though some of the UMTS900 cells near the railway track will have more capacity loss than other cells. In the case where GSM-R uplink power control is not used, the simulation results show that much more capacity loss on UMTS UL can occur, especially for the cells located near the railway track Conclusions Based on the co-existence scenarios between UMTS900 and GSM-R, the simulation assumptions described in section 3.2.4, and the simulation results and analysis on GSM-R DL outage probability and UMTS UL capacity loss, the following conclusions can be made: UMTS900 can be deployed in the same geographical area in co-existence with GSM-R as follows: 1) There is a priori no need of an additional guard band between UMTS900 and GSM-R, a carrier separation of 2.8 MHz or more between the UMTS900 carrier and the nearest GSM-R carrier is sufficient without prejudice to provisions in point 2). This conclusion is based on Monte Carlo simulations assumed suitable for typical case. 2) However for some critical cases (e.g. with high located antenna, open and sparsely populated areas served by high power UMTS BS close to the railway tracks, blocking etc, which would lead to assumption of possible direct line of sight coupling) the MCL calculations demonstrate that coordination is needed for a certain range of distances (up to 4 km or more from railway track). 3) It is beneficial to activate GSM-R uplink power control, especially for the train mounted MS, otherwise the impact on UMTS UL capacity could be important when the UMTS network is using the 5 MHz channel adjacent to the GSM-

24 Page 24 R band. However, it has to be recognized that this is only applicable in low speed areas as elsewhere the use of uplink control in GSM-R will cause significantly increased call drop out rates. 4) In order to protect GSM-R operations, UMTS operators should take care when deploying UMTS in the 900 MHz band, where site engineering measures and/or better* filtering capabilities (providing additional coupling loss in order to match the requirements defined for the critical/specific cases) may be needed in order to install UMTS sites close to the railway track when the UMTS network is using the 5 MHz channel adjacent to the GSM-R band. * Currently, the out-of band interference level is given by 3GPP TS V7.4.0 It has to be noted that this study did not address tunnel coverage. Site sharing, which is expected to improve the coexistence, has not been studied either. 3.3 Compatibility consideration between UMTS900 and PMR/PAMR Characteristics of PMR/PAMR systems Several radio systems will potentially use the PMR/PAMR frequency band, such as TETRA, CDMA PAMR, TAPS, etc CDMA PAMR system characteristics The system description of CDMA PAMR can be found in ETSI harmonized standard EN for CDMA PAMR [8]. The main CDMA PAMR system characteristics are summarized in tables 3-10 to CDMA PAMR Frequency band (UL) (MHz) Frequency band (DL) (MHz) Carrier separation (MHz) 1.25 Modulation BS-MS MCL (db) BS QPSK/BPSK 70 (Urban area) 80 (Rural area) Maximum Tx power (dbm) Thermal noise (dbm) Noise figure (db) 5 9 Noise floor (dbm) Receiver sensitivity (dbm) Antenna height (m) 30 (Urban) (Rural) Antenna gain (dbi) 17 0 Feeder loss (db) 2 0 ACS (db) Table 3-10: Main CDMA PAMR system parameters MS

25 Page 25 For f Within the Range Applicability Emission Limit 750 to 885 KHz Single Carrier ( f -750)/135 dbc in 30 khz 885 to 1125 KHz Single Carrier -60-5( f -885)/240 dbc in 30 khz to 1.98 MHz Single Carrier -65 dbc / 30kHz 1.98 to 4.00 MHz Single Carrier -75 dbc / 30kHz 4.00 to 6.00 MHz Single and Multiple Carrier -36 dbm / 100kHz 6.00 to MHz Single and Multiple Carrier -45 dbm / 100kHz > MHz Single and Multiple Carrier -36 dbm / 1 khz; -36 dbm / 10 khz; -36 dbm / 100 khz -30 dbm / 1 MHz; 9 khz < f < 150 khz 150 khz < f < 30 MHz 30 MHz < f < 1 GHz 1 GHz < f < 12.5 GHz Table 3-11: CDMA PAMR BS spectrum mask (Transmitter unwanted emission limits for Band Class 12) For f within the range f Within the Range Applicability Emission Limit 1.98 to 4.00 MHz Single Carrier -100 dbc / 30kHz 4.00 to 6.00 MHz Single and Multiple Carrier -61 dbm / 100kHz >6.00 MHz Single and Multiple Carrier -61 dbm / 100kHz Table 3-12: Additional BS Transmitter unwanted emission limits for Band Class 12 within the frequency range MHz For f Within the Range Emission Limit 885 khz to MHz ( f 885) / 235 dbc in 30 khz MHz to 1.98 MHz ( f 1120) / 860 dbc in 30 khz 1.98 MHz to 4.00 MHz ( f 1980) / 2020 dbc in 30 khz 4.00 MHz to 10.0 MHz -51 dbm in 100 khz >10.0 MHz -36 dbm/1 khz; -36 dbm/10 khz;-36 dbm/100 khz;-30 dbm/1 MHz; 9 khz < f < 150 khz150 khz < f < 30 MHz30 MHz < f < 1 GHz1 GHz < f < 12,75 GHz Table 3-13: CDMA PAMR MS Spectrum mask (Unwanted emission limits for mobile stations) Frequency Maximum E.R.P/ reference bandwidth 30 MHz f < MHz -36 dbm/100 khz 1 GHz f < 12,75 GHz -30 dbm/1 MHz Fc1-4 MHz < f < Fc2 + 4 MHz No requirement NOTE 1: Centre frequency of first carrier frequency (Fc1) used by the base station. NOTE 2: Centre frequency of last carrier frequency (Fc2) used by the base station. NOTE 3: Note 1 and Note 2 assume contiguous frequencies otherwise multiple exclusion bands will apply. Table 3-14: BS Spurious emission (Radiated unwanted emissions requirements)

26 Page 26 Frequency Limit (E.R.P)/ reference bandwidth idle mode Limit (E.R.P)/ reference bandwidth traffic mode 30 MHz f < MHz -57 dbm/100 khz -36 dbm/100 khz 1 GHz f < 12,75 GHz -47 dbm/1 MHz -30 dbm/1 MHz Fc - 4 MHz < f < fc + 4 MHz No requirement No requirement NOTE: fc is the nominal MS transmit centre frequency. Table 3-15: MS Radiated unwanted emissions requirements TETRA system characteristics The main TETRA system characteristics are summarized in tables 3-16 to TETRA Frequency band (UL) (MHz) Frequency band (DL) (MHz) Carrier separation (MHz) BS-MS MCL (db) BS 25 khz 70 (Urban area) 80 (Rural area) Maximum Tx power (dbm) Receiver bandwidth (khz) Thermal noise (dbm) Noise figure (db) 5 9 Noise floor (dbm) Receiver sensitivity (dbm) Antenna height (m) 30 (Urban) (Rural) Antenna gain (dbi) 14 0 Feeder loss (db) 2 0 Receiver protection ratio (db) Table 3-16: Main TETRA system parameters MS Frequency Offset 30 dbm Mobile Station 44 dbm Base Station 25 khz - 30 dbm - 16 dbm 50 khz -40 dbm - 26 dbm 75 khz -40 dbm - 26 dbm khz -45 dbm - 36 dbm khz -50 dbm - 41 dbm 500 khz - f rb - 50 dbm - 46 dbm > f rb - 70 dbm - 56 dbm Table 3-17: TETRA Spectrum Mask* *measured in an 18 khz bandwidth. *f rb is the edge of the receive band belonging to the TETRA MS/BS. The minimum unwanted emissions requirement is - 36 dbm for frequency offsets of 25, 50 and 75 khz and - 70 dbm for higher offsets.

27 Page 27 Frequency Offset MS BS khz - 40 dbm -40 dbm khz - 35 dbm - 35 dbm khz - 30 dbm - 30 dbm > 500 khz - 25 dbm - 25 dbm Table 3-18: TETRA Receiver Blocking Interference analysis considerations It can be seen that the UMTS900 UL frequency block ( MHz) is adjacent to the PMR/PAMR system (CDMA PAMR or TETRA) DL frequency block ( MHz) at the frequency 915 MHz. The worst interference scenario between UMTS900 uplink and PMR/PAMR system downlink (CDMA PAMR or TETRA) could potentially happen at around 915 MHz. ECC Report 82 (section ) [1] indicated that UMTS outdoor UE transmitting power is relatively small, at 90% percentile, the simulated outdoor UE transmit power is dbm. By considering that the minimum coupling loss between UE and PMR/PAMR BS is relatively large (80 db is used in ECC Report 82 between UE and BS in rural area) compared to the MCL between UMTS BS and GSM-R Train Mounted MS, and since the UE is moving, the interference from UMTS UE to PMR/PAMR BS should not be a problem. For detailed analysis of interference between UMTS UE and PMR/PAMR BS, Monte-Carlo simulations should be performed; this is not covered in this report. The worst interference case is the interference from PMR/PAMR BS to UMTS BS, as shown in figure Potential interference between UMTS900 and CDMA PAMR at 915 MHz Interference from CDMA PAMR BS operating between MHz to GSM900 BS operating below 915 MHz with a frequency separation of 2.15 MHz was analyzed in ECC Report 41 [7]. As described in ECC Report 41 [7], a frequency separation of 2.15 MHz between GSM900 operating below 915 MHz and CDMA PAMR operating above 917 MHz is not sufficient for the protection of GSM900 BS receiver; coordination between GSM900 and CDMA PAMR is recommended in ECC Report 41 [7]. CDMA-PAMR900 UMTS900 Figure 3-13: Worst Interference scenario between CDMA PAMR downlink and UMTS900 uplink As shown in figure 3-13, the potential interference from CDMA-PAMR BS can desensitize UMTS900 BS receiver if the protection is not sufficient. UMTS900 system parameters are described in ECC Report 82[1]. The interference protection level for UMTS900 BS receiver is -110 dbm/3.84 MHz and the ACS of UMTS900 BS receiver is 46.2 db. Based on the CDMA PAMR BS spectrum mask for band class 12 given in tables 3-11 and 3-12, for a guard band of 0.6 MHz between a UMTS900 carrier below 915 MHz and a CDMA PAMR carrier above 915 MHz, the required MCL between UMTS900 BS and CDMA PAMR BS is 95.6 db. When using a free space propagation model, the space

28 Page 28 separation between UMTS900 BS and CDMA PAMR BS antennas is in the order of 8 km. However, when using the Hata propagation model, the separation distance becomes 1.5 km MCL , Guard band (MHz) Figure 3-14: Required MCL (db) in function of guard band The required MCL as function of guard band is given in figure It indicates the required MCL decreases when the guard band becomes larger. Two possible solutions can be used to meet the required MCL between UMTS900 BS and CDMA PAMR BS: a) Space separation; b) external filter Potential interference between UMTS900 and TETRA at 915 MHz The adjacent compatibility study between GSM900 and TETRA or TAPS at 915 MHz was described in ECC Report 5 [9] showing that without any guard band or other interference mitigation techniques, interference from TETRA/TAPS BS will desensitize GSM900 BS receivers. In order to protect the GSM900 BS receiver operating below 915 MHz, several interference mitigation techniques were recommended in ECC Report 5 for the protection of GSM900 BS receivers, such as guard band, filters, and/or coordination between operators. The interference analysis method described in ECC Report 5 can be re-used for the interference analysis between UMTS900 and TETRA systems operating below and above 915 MHz respectively, by considering that UMTS900 BS is more sensitive to interference than GSM900, the maximum tolerable interference level for the protection of UMTS BS receiver is of -110 dbm/3.84 MHz. By applying the interference analysis method described in ECC Report 5, similar conclusions can be made that without interference mitigation techniques there will be serious interference from a TETRA/TAPS BS transmitter to UMTS900 BS. Thus UMTS900 BS receivers will be desensitized due to strong interference from TETRA/TAPS. The following interference mitigation techniques can be used to reduce the interference from TETRA/TAPS to UMTS900 BS: i) Guard band; ii) External filters; iii) Spatial separation by coordination between UMTS900 and TETRA/TAPS operators; iv) Reduced transmitting power of TETRA/TAPS BS Conclusions The interference from PMR/PAMR (CDMA PAMR, TETRA, TAPS) BS operating at frequencies above 915 MHz will cause receiver desensitization of UMTS900 BS operating below 915 MHz. In order to protect UMTS900 BS, the utilization of interference mitigation techniques is necessary: 1) Reduced PMR/PAMR BS Tx power; 2) Spatial separation by coordination between operators; 3) External filters; 4) Guard band. It is more likely that a combination of these interference mitigation techniques should be used in order to ensure the compatibility between UMTS900 operating below 915 MHz and PMR/PAMR (CDMA PAMR, TETRA, TAPS) operating above 915 MHz.

29 Page Compatibility study between UMTS900 and DME DME and UMTS system characteristics Protection criteria for the aeronautical radionavigation service The protection criteria for the aeronautical radionavigation service are extracted from Recommendation ITU-R M Recommendation ITU-R M.1639 gives the equivalent power flux-density (EPFD) level which protects stations of the aeronautical radionavigation service (ARNS) from emissions of radionavigation satellites of all radionavigation-satellite service (RNSS) systems operating in the MHz band. It recommends that the maximum allowable epfd level from all space stations of all RNSS systems should not exceed db(w/(m 2 MHz)), in order to protect the ARNS in the band MHz. The instantaneous epfd is calculated using the following formula: Na Pi G θ ϕ = t ( i ) G r ( i ) epfd 10 log (Equation 1) 10 2 i= πdi G 1 4 r, max where: N a : i : P i : θ i : G t (θ i ) : d i : ϕ i : G r (ϕ i ) : G r,max : epfd : number of space stations that are visible from the receiver index of the space station considered RF power at the input of the antenna (or RF radiated power in the case of an active antenna) of the transmitting space station (db(w/mhz)) off-axis angle between the boresight of the transmitting space station and the direction of the receiver transmit antenna gain (as a ratio) of the space station in the direction of the receiver distance (m) between the transmitting station and the receiver off-axis angle between the pointing direction of the receiver and the direction of the transmitting space station receive antenna gain (as a ratio) of the receiver, in the direction of the transmitting space station (see Recommendation ITU-R M.1480) maximum gain (as a ratio) of the receiver instantaneous epfd (db(w/(m 2 MHz))) at the receiver

30 Page 30 The maximum allowable aggregated EPFD levels for protecting ARNS are summarized in table Parameter Value Reference DME RNSS interference threshold (at antenna port) Maximum DME/TACAN antenna gain including polarization mismatch Effective area of 0 dbi antenna at MHz RNSS (all systems) aggregate epfd in 1 MHz 129 db(w/mhz) (See Note 1) 3.4 dbi 22.9 db(m 2 ) db(w/(m 2 MHz)) 5 Safety margin 6 db 6 Apportionment of RNSS interference to all the interference sources 6 db (5.4 dbi antenna gain 2 db polarization mismatch) Combine 1, 2 and 3 (1 minus 2 minus 3) Recommendation ITU-R M.1477 Apportion 25% of total permissible interference to RNSS 7 Maximum RNSS aggregate epfd db(w/(m 2 Combine 4, 5 and 6 MHz)) (4 minus 5 minus 6) Table 3-20: maximum allowable aggregated EPFD level to protect ARNS from RNSS NOTE 1 This value is based on a 129 dbw CW interference threshold specified for international DME systems used by civil aviation. Measurement has demonstrated that an RNSS signal spread over 1 MHz would have the same effect as a CW signal on DME performance. Transposition to UMTS 900 A more convenient way to convert the above criteria to UMTS 900 is to express it as a PSD received at the DME antenna port, including the safety margin and the apportionment, as given in table Parameter Value Reference DME interference threshold (at DME antenna port) 129 db(w/mhz) 2 Safety margin 6 db 3 4 Apportionment of UMTS interference to all the interference sources (MIDS, FRS, etc.) Maximum UMTS aggregate PSD, received at the DME receiver input, including the safety margin and the apportionment 6 db 141 db(w/mhz) Recommendation ITU-R M.1477 Apportion 25% of total permissible interference to UMTS. It is noted that higher percentage could be considered in the band MHz. Combine 1, 2 and 3 (1 minus 2 minus 3) Table 3-21: Maximum allowable aggregated PSD level to protect ARNS from UMTS900 The following aggregated PSD value must not exceed -141 db(w/mhz): PSD Na = log i= 1 Pi λ G t ( θ i ) Gr ( ϕi ) (Equation 2) 4πd i

31 Page 31 where : N a : i : P i : λ : θ i : G t (θ i ) : d i : ϕ i : G r (ϕ i ) : PSD : number of UMTS 900 base stations that are visible from the receiver (DME) index of the base station considered RF power at the input of the antenna the transmitting UMTS 900 base station (db(w/mhz)) wave length off-axis angle between the boresight of the transmitting UMTS 900 base station and the direction of the receiver (DME) transmit antenna gain of the base station in the direction of the receiver (DME) distance (m) between the transmitting base station and the receiver off-axis angle between the pointing direction of the receiver and the direction of the transmitting UMTS 900 base station receive antenna gain of the receiver (DME), in the direction of the transmitting UMTS 900 base station instantaneous PSD (db(w/(mhz))) at the receiver (DME) It has to be noted that the threshold above was established by measurement of a number of DME airborne receivers (interrogator receiver) under various signal conditions and confirmed that the effect of an RNSS signal, when spread over 1 MHz, had the same effect on the DME receiver as does CW. As the DME specification requires correct performance in the presence of CW at -129 db(w/mhz), this was given as the appropriate maximum level for all RNSS interference. The same assumption was made when modelling the effect of the interference from UMTS900 on DME. This is justified by the nature of the UMTS900 signal (W-CDMA spread signal). Set of DME parameters Frequency of band of operation: MHz Receiving frequency (in the simulation) : 962, 964, 966 and 971 MHz Polarization: linear, vertical (so no polarization loss should be considered) Maximum DME antenna gain : 5.4 dbi Channelization: 1 MHz Bandwidth : 1 MHz ARNS station location: the ARNS station altitude should be taken at worst case ( ft, i.e m), which gives maximum visibility of potentially interfering base stations from the ARNS receiving antenna. DME Selectivity mask: o DME 442 Rockwell Collins. The attenuations are 6 db at MHz/+0.32 MHz (-0.88 MHz/+0.82 MHz from the central frequency) 20 db at MHz/+0.49 MHz (-1.05 MHz/+0.99 MHz from the central frequency) 40 db at MHz/+0.62 MHz (-1.30 MHz/+1.12 MHz from the central frequency) 60 db at MHz/+0.64 MHz (-1.46 MHz/+1.14 MHz from the central frequency) o KN 62A Honeywell. The attenuations are 6 db at MHz/+0.34 MHz (-0.65 MHz/+0.84 MHz from the central frequency) 20 db at MHz/+0.48 MHz (-0.76 MHz/+0.98 MHz from the central frequency) 40 db at MHz/+0.49 MHz (-0.79 MHz/+0.99 MHz from the central frequency) 60 db at MHz/+0.50 MHz (-0.80 MHz/+1.00 MHz from the central frequency) It has to be noted that the values of the selectivity masks have set to 70 dbc beyond 250% of the bandwidth (+/- 2.5 MHz) with a linear interpolation between 60 and 70 dbc.

32 Page 32 ARNS antenna characteristics The information in the following Fig. 15(a).a is extracted from Recommendation ITU-R M.1642 and provides the antenna gain for different elevation angles. For intermediate elevation angles (between two defined values), a linear interpolation should be used. The G r, max value is 5.4 dbi as specified in Recommendation ITU-R M It is assumed that the elevation and gain pattern is the same for all azimuth angles. The relevant range of elevation angles for the study to be conducted is: -90 0, as shown in Fig. 15(a). Elevation angle (degrees) Extract from Rec. ITU-R M.1642 Antenna gain G r /G r, max (db) Elevation angle definition Figure 3-15(a): DME antenna gain for elevation angles between 0-90 Proposed parameters for UMTS 900 Therefore, the scenario worth studying is the situation where multiple base stations produce interference to onboard DME: Antenna input Power: 43 dbm/channel (for Macro base stations). Micro and pico base stations have not been considered. It has to be noted that this figure represents a fully loaded cell. Average cell radius : 5 km Unwanted emissions characteristics : see Table 3.22 below Channel Spacing : 5 MHz Maximum antenna gain including the feeder loss : 15 dbi Receiver Bandwidth : 3840 KHz Elevation antenna pattern : Recommendation ITU-R F Azimuth antenna pattern : omni-directional Downtilt : 2.5 Antenna height : 30 m For information, a comparison between the out-of-band emissions of UMTS-900 and GSM 900 is available in section

33 Page 33 Out of band domain Spurious domain Frequency offset to the UMTS central frequency 2.5 MHz f < 2.7 MHz 2.7 MHz f < 3.5 MHz Power density (cabinet output) Power in 5 MHz Level in dbc -14 dbm/30 khz 8 dbm -35 dbc Linear interpolation Linear interpolation Linear interpolation f = dbm/30 khz -4 dbm -47 dbc 3.5 MHz < f -13 dbm/1 MHz -6 dbm -49 dbc 12.5 MHz f =12.5 MHz to -36 dbm/100 khz -19 dbm -62 dbc frequency=1 GHz Frequency>1 GHz -30 dbm/1 MHz -23 dbm -66 dbc Table 3-22: Unwanted emissions characteristics for UMTS900 Common parameters Frequency plan is given in table MHz 915 MHz 925 MHz 960 MHz UMTS-UL UMTS-DL DME Table 3-23: frequency plans Propagation model: Free space loss (Recommendation ITU-R P.525) : all the base stations are visible from the aircraft, without any obstacle Case Study The interference on the DME comes from all the base stations which have visibility of the aircraft at its altitude, see Fig. 3-15(b). Considering a frequency re-use scheme of 1, each base station transmits 3 carriers at full power. The base stations generate 3 sub-interferences at the following frequencies: f 1 =957.5 MHz (1 st adjacent channel interference to be considered) f 2 =952.5 MHz (2 nd adjacent channel interference to be considered) f 3 =947.5 MHz (3 rd adjacent channel interference to be considered) In practise, each UMTS900 base station may transmit more than 3 carriers but the other ones are not considered in these simulations.

34 Page 34 DME BTS : f 1, f 2, f 3 BTS BTS : f 1, f: 2, f 1, f 3 f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 BTS : f 1, f 2, f 3 Figure 3-15(b): Scenario of the study The principles are: To distribute the base stations on the terrestrial dome seen by the DME (placed every 10 km); To assess the aggregated signal generated by signals from the base stations at f 1, f 2 and f 3 ; To compare this aggregated signal to the threshold of -141 db(w/mhz) Interference analysis results Calculation of the ACIR The UMTS900 ACLR and DME ACS are plotted in figure 3-16(a), ACIR of DME as function of frequency separation and as guard band are respectively given in figures 3-16 (b) and (c).

35 Page 35 (a) UMTS900 ACLR and DME ACS (b) ACIR as function of frequency separation

36 Page 36 (c) ACIR as function of guard band Figure 3-16: Calculation of the ACIR Number of visible base stations The number of visible base stations as a function of aircraft altitude is given in figure Figure 3-17: Number of visible base stations Calculation of the UMTS aggregate PSD The calculated UMTS aggregate PSD and corresponding margin to add for satisfying the interference criteria for three DME frequencies are given in Fig It has to be noted that the two DME equipments mentioned previously have been considered in the calculations. Additionally, the results inherent to an ideal DME filter are given in the following table for information.

37 Page 37 DME freq calculated UMTS aggregate PSD additional margin needed 968 MHz 970 MHz 972 MHz Figure 3-18: Calculated UMTS aggregate PSD and corresponding margin to add to satisfy the interference criteria

38 Page Analysis of the results The selectivity of two DME airborne receivers has been measured. However, the characteristics of the receiving filter are only given for the range MHz from the central frequency. One can easily assume that the filter continues to decrease after this value but in the absence of data, the filter has been considered as flat after these values. The parameters of the UMTS equipment are based on the 3GPP standards, except for the base station antenna pattern which is based on the Recommendation ITU-R F The results depend on the altitude of the aircraft. There is a difference of around 3 db between the low altitudes (100m<Altitude<500m and the cruise situations (12000 m). The results are nearly the same for the two DME equipments considered. Given the data which have been taken into account (and expressed in the previous paragraphs), some additional isolation (to satisfy the interference criteria) may be needed to make the compatibility between UMTS900 and DME possible, and particularly if the DME frequency is below 972 MHz. It has to be noted that the value of this additional isolation depends on the value of the DME carrier: Flight phase DME 442 Rockwell Collins and DME KN 62A Honeywell DME carrier 962 MHz 964 MHz 966 MHz 968 MHz 970 MHz 972 MHz 0 m<altitude<100m m<altitude<500m m<altitude<2000 m Cruise Table 3-24: additional margin needed (db) It has to be noted that certain factors or parameters, relating to the deployment and/or the definition of the UMTS900 system are not stabilized yet. These factors/parameters are: A traffic load factor: The global load is referring to the distribution of traffic load within the UMTS network. Even when considering the peak of traffic, a minority of UMTS BS is fully loaded and is transmitting at Pmax. The operators and manufacturers have confirmed the consideration of a traffic load factor for the design of their networks. The maximum value for a loaded cell is assumed to be 80%, whereas an average value is 50%. It is recognized that a safety aeronautical system such as DME has to examine the worst case in terms of interference for UMTS900, when the number of UMTS900 cells considered is low (i.e. when the altitude of the aircraft is low). However, when the altitude of the aircraft is higher, it is reasonable to consider the average load factor value (because the number of the base stations considered is high). Given the number of visible base stations as shown in the following figure:

39 Page 39 it may be assumed that the average value can be applied from an altitude of 100m (150 base stations seen by an aircraft): Value of the traffic load factor Altitude of the aircraft <100m 80% Altitude of the aircraft >100m 50% Maximum antenna gain of the UMTS900 base stations: The value of 15 dbi (including feeder loss) has been considered which corresponds to the rural case. However, it is recognized that the antennas deployed in urban areas have commonly a gain of 12 dbi (including feeder loss). The ratio of one antenna type compared to the other one is currently not defined. It has also to be noted that if certain parameters are adjusted according to the previous bullets, those adjustments have to be consistent (E.g : a rural case could correspond to an antenna gain of 15 dbi including feeder loss associated to a low traffic load). It should be emphasized that there may be a need for additional calculations to model the approach phase (or other phase) when an aircraft rolls. In this configuration, the maximum gain of the DME (reception) corresponds to an elevation angle of -20/-25 (see table 3-22). It should also be noted that no compatibility study between GSM900 and systems operating in adjacent band had been performed prior to the deployment of GSM900. This has not been represented a problem so far since the aeronautical equipments do not currently use the part of the band just above 960 MHz: The lowest frequency used by DME is 977 MHz; The lowest frequency used by TACAN is 978 MHz. According to the recognized international aviation standards, the frequency range for the DME is MHz and carriers lower than 977 MHz, such as 962 MHz, may also be deployed in the future. In any case, the use of TACAN/DME below 970 MHz requires additional protection so that the compatibility in adjacent band with UMTS900 can be ensured. Moreover, it has to be noted that the frequencies just above 960 MHz are also under consideration under the AI 1.6 of the next WRC for the development of new aeronautical mobile systems in that band Mitigation techniques and mitigation effects Mitigation techniques and mitigation effects are therefore required, such as: The reduction of the out-of-band UMTS900 emission: this is achieved with the use of UMTS900 base stations with out-of-band performances better than the requirement defined in the 3GPP specifications (e.g.: filtering); this may not be technically feasible to ensure the protection of all DME frequencies. (e.g.: 962 to 966 MHz); Site engineering for the UMTS 900 base stations situated in/near the airports to achieve additional protection for the takeoff/landing phases ; this can be implemented only on a limited number of base stations (which depends on the nature of the specific site engineering measures); Consideration of a sufficient guard band, considering that there is already a 1.5 MHz guard band ( MHz); Examination of lower apportionment margin: it has to be noted that the military MIDS system does not operate in the lower part of the band ( MHz); therefore, the apportionment margin can be reduced. This reduction has not been considered in the above calculations. The appropriate value of the apportionment is 3 db if the interferences to DME are assumed to equally come from UMTS900 and the potential FRS system (Futur Radio System) considered under the agenda item 1.6 of the WRC-07. This is subject to the result of the WRC-07. It has also to be noted that the FRS system is not likely to be deployed before Therefore the apportionment should be alleviated as follows:

40 Page 40 Before the deployment of the FRS system (before 2020) After the deployment of the FRS system (before 2020) MHz Apportionment = 0 db Apportionment = 3 db In the upper part of the band ( MHz), the interferences from the MIDS have also to be considered. However, MIDS is a frequency hopping system that hopes on 51 frequencies, the first ones of which are 969, 972 and it is recognized that the interferences from UMTS900 above 972 are negligible. The value of the apportionment is calculated as follows: 2/51*X (MIDS) +X (UMTS900) =1 (before 2020) X=0.96 %= 0.16 db 2/51*X (MIDS) +X (FRS) +X (UMTS900) =1 X=0.49%= 3.1 db Which gives: Before the deployment of the FRS system (before 2020) After the deployment of the FRS system (before 2020) MHz Apportionment = 0.16 db Apportionment = 3.1 db This leads to the conclusion that the interferences from MIDS have a negligible effect Conclusions Under the assumptions described above, the following preliminary conclusions can be made based on simulation results and interference analysis: Nowadays, the lowest DME frequency is 977 MHz. Lower frequencies are planned to be used for DME in a near future; As long as the DME frequencies are above 972 MHz, the electromagnetic compatibility between DME and UMTS 900 is ensured without any care to be taken; Regarding the frequencies from 960 to 972 MHz, the only mitigation techniques, in order to ensure the compatibility between the DME system and UMTS900, that would bring sufficient isolation are: additional filtering and a larger guard band. However these two mitigation techniques are not judged applicable for the following reasons: o Additional filtering: the UMTS900 manufacturers have clearly indicated that, nowadays, it is not technologically feasible to provide the sufficient margin needed (compared to the specified out-of-band emission mask considered in the above calculations) without affecting the level of the transmitted power in the transmitting band (it is recognized that the introduction of additional filtering creates insertion losses of several dbs on base stations transmission power level that need to be balanced by increasing deployment density); o Larger guard band: the above calculations have shown that an additional 10 MHz guard band (to the existing 1.5 MHz guard band) is needed. This is unacceptable for both UMTS900 and civil aviation communities; There is a need for consideration of this issue on a European context, on the regulatory aspect. It is necessary that a common approach be used within Europe to ensure the compatibility. It has to be noted that the impact of the DME ground station (and FRS if necessary) on the UMTS 900 mobile stations has not been studied in this report and may need additional studies.

41 Page Compatibility study between UMTS900 and MIDS System parameters and co-existence scenario Frequency band plan The frequency band plans for MIDS and UMTS900 are shown in figure below: UMTS900 Mobile station transmit UMTS900 Base station transmit MIDS Terminals receive UMTS900 frequency band is arranged as follows: Uplink (UE transmit, BS receive) : MHz Downlink (BS transmit, UE receive) : MHz Carrier separation: 5 MHz MIDS frequency band is arranged as follows: MIDS operates in the 960 to 1215 MHz band, with MIDS frequencies occurring every 3 MHz between 969 to 1206 MHz. Two sub-bands centered on 1030 MHz and 1090 MHz are excluded because they are used by IFF System parameters UMTS900 parameters The characteristics of UMTS900 system are summarized in the table3.25 below. Parameter Carrier spacing Duplex method IMT-2000 CDMA Direct Spread (UMTS900) 5 MHz ± n 0.2 MHz FDD Cell type Macro Micro Pico Transmitter power dbm (3) Antenna gain (4), (5) (dbi/120 sector) 15 (6) 5 0 Cable loss Antenna height (m) Tilt of antenna (degrees down) Access techniques Data rates supported Modulation type Emission bandwidth Transmitter ACLR for macro/micro/ pico BS 1st adjacent 2nd adjacent Transmitter spurious emissions Receiver blocking levels CDMA Pedestrian: 384 kbit/s, Vehicular: 144 kbit/s, Indoors: 2 Mbit/s Higher data rates up to 10 Mbit/s are supported by technology enhancements (HSDPA) QPSK 3GPP TS GPP TS ± 5 MHz 50 ± 10 MHz 3GPP TS GPP TS Table 3.25: UMTS900 base station parameters

42 Page 42 (3) (4) (5) (6) May not be appropriate for all scenarios. Feeder losses are not included in the values and should be considered in the sharing/compatibility issues. The reference pattern is specified in Recommendation ITU-R F.1336 with (k = 0.2). Cable loss is included in the antenna gain. In order to have a realistic representation of UMTS900 equipments, two types of cells have been considered: macro and micro: Macro Cell Micro Cell Transmission power +43 dbm (20W) +38dBm (6,3W) Cable loss 3dB (included in antenna gain) 1dB Antenna gain 15dBi (120 sector) 5dBi (omni-directional or directive) PIRE 58dBm 42dBm Antenna height 40m 5m Vertical aperture 6 Downtilt 2,5 0 For micro cell and macro cells, unwanted emission limits in the out-of band domain and in the spurious domain are defined below: Out of band domain Spurious domain (in accordance with ITU-R SM329) F LIMIT 1 st adjacent channel ± 5 MHz 45 dbc 2 nd adjacent channel ± 10 MHz 50 dbc Between 960MHz and 1 GHz Between 1GHz and 12 GHz -36dBm in 100kHz or-19 dbm in 5 MHz -30dBm in 1MHz or -23 dbm in 5 MHz Other assumptions taken in the study: Emission frequency of the base station: MHz (highest UMTS900 channel between 955 and 960 MHz); It is assumed that the UMTS900 antenna has no attenuation in the receiving band of MIDS ( MHz). This represents a worst case; The UMTS900 base station transmits continuously MIDS parameters MIDS (Multifunctional Information Distribution System) is a tactical military system. The MIDS receiver to consider is the MIDS terminal, integrated in a shelter. The antenna is mounted on a 16 metres mast. The terminal mode to consider is the frequency hopping mode (51 frequencies). The lowest frequency is 969 MHz.

43 Page 43 Receiver MIDS terminal Bandwith 5 MHz Feeder loss 5 db Antenna gain 9 dbi Antenna height 16 metres Equivalent downtilt db beam width in the 16 vertical plane Horizontal plan omni Table 3.26: MIDS terminal parameters 100 MIDS terminal antenna Gain (in 60 %) elevation (degre) 1000 MHz 1100 MHz 1200 MHz Figure 3-19: MIDS terminal elevation antenna diagram

44 Page 44 Other assumptions taken in the study: The frequency used in the simulation is 970 MHz; The hopping frequencies are as follows: N Frequency (MHz) N Frequency (MHz) N Frequency (MHz) Protection from the unwanted emissions of an interfering system: criterion n 1 Measurements have been performed in a French DoD laboratory to assess the protection criteria of MIDS receiver. The curves of the permissible level of a signal which is out of the MIDS band, have been picked out: for a transmission at 960 MHz, there is no degradation of the MIDS terminal performances as long as the power of the transmitter remains below -10 dbm (the reference is a CW signal). Noise level permissible in the MIDS channel: criterion n 2 On the same line as in the previous paragraph, measurements on MIDS receiver give a permissible noise level equal to -103dBm, for one of the 51 channels, i.e dbm/5 MHz, taking into account 1dB margin This tolerated value allows to obtain an acceptable MIDS sensitivity referred to MIDS SSS (System Segment Specification). Interference threshold expressed as an interfered frequencies rate The MIDS receiver can tolerate a certain number of interfered channels amongst the 51 channels used, without any performance degradation. This threshold is classified and is not given in this document. This interference threshold, without being communicated in the report for security reason, is covered by the criterion n 2, when assessing the number of frequencies for which the permissible noise floor is exceeded.

45 Page Propagation model Propagation model used is ITS. This model is usually used for MIDS studies, in France as well as in USA (NTIA) Simulation configuration The simulation configuration is given in the figure 3-20 below. UMTS base station antenna MIDS terminal Antenna height Distance between the MIDS terminal and the UMTS base station Figure 3-20: Illustration of the simulation configuration Interference analysis and simulation results Level of the UMTS900 signal received by the MIDS terminal (out of the MIDS receiving band) The aim of this section is to assess the interference from the UMTS900 base station in the UMTS900 band, according the criterion n 1 described above.

46 Page Simulation results Macro cell The following curves give the level of the UMTS900 signal as a function of the distance between the UMTS900 base station and the MIDS terminal: Figure 3-21: MACRO base station transmitted power received by the MIDS terminal (distance<1km) Figure 3-22: MACRO base station transmitted power received by the MIDS terminal (distance<5km) Analysis: Whatever the distance between the UMTS900 base station and the MIDS terminal is, the maximum authorized level of -10 dbm is not exceeded. Maximum level equal to -21dBm is reached for a distance from 200m to 280m. Micro cell The following curves give the level of the UMTS900 signal as a function of the distance between the UMTS900

47 Page 47 Figure 3-23: MICRO base station transmitted power received by the MIDS terminal (distance<1km) Analysis: The level of -10 dbm is never reached. Maximum level equal to -38dBm is reached for a distance of 160m Conclusion according to criterion n 1 Considering the protection curves of MIDS receiver, there is no risk of saturation caused by UMTS signal Level of the UMTS900 signal received by the MIDS terminal (in the MIDS receiving band) The aim of this section is to assess the interference from the UMTS900 base station in the UMTS900 band, according the criterion n 2 described above. The out-of band and spurious emissions of the UMTS900 are considered in this section UMTS900 BS unwanted emissions UMTS900 Macro-cell (Power = 43 dbm, e.i.r.p. = 58dBm) UMTS900 Transmission band Level in dbc in 5 MHz e.i.r.p. in dbm in 5 MHz MIDS channels impacted MHz 50 dbc +8 dbm 969 MHz (1 channel) 970 MHz to 1 GHz (62dBc) -4 dbm 972 to 999 MHz (10 channels) 1 GHz to 12,75 GHz (66dBc) -8 dbm 1002 to 1206MHz (40 channels) Figure 3-24: UMTS900 BS unwanted emission (macro cell)